EP0407879B1 - Method for mutli-dimensional measurement of magnetic resonance in defined small volume regions of a solid-state sample - Google Patents

Description

The invention relates to a method for multidimensional measurement of magnetic resonance in defined small volume areas of a solid sample, in which the sample is arranged in a constant, homogeneous magnetic field and irradiated in a predetermined manner with a sequence of high-frequency pulses and a sequence of gradient magnetic field pulses is suspended in such a way that the magnetization of the spins to be measured is maintained for a period of time which is longer than a switch-off period of the gradient magnetic field pulses.

A method of the type mentioned above is known from US-Z Journal of Magnetic Resonance 66 (1986), pages 530 to 535.

It is generally known to use the technique of magnetic resonance, in particular nuclear magnetic resonance, for two-dimensional or three-dimensional measurement in defined small volume ranges of samples. In medical research and medical diagnosis in particular, measurements are carried out on living or non-living human tissue in this way. A distinction is made between localized spectroscopic measurements, in which a nuclear magnetic resonance spectrum is only recorded in a small volume range, and, on the other hand, between imaging methods, in which a two- or three-dimensional representation of a body part is recorded as an image of the spin density or relaxation time.

Up to now, such volume-selective measurements have mainly been carried out on liquid samples. This is possible in a simple manner because liquid samples have relatively narrow lines or slow decay times of the pulsed nuclear magnetic resonance signals due to the molecular movement and the necessary measuring steps can be carried out during the relatively long duration of these signals in the order of magnitude of several 100 ms in order to achieve a small one Select and read volume range. These measuring steps consist in essentially in the application of gradient magnetic field pulses, for which a certain minimum switch-on time and a minimum switch-off time are required for technical reasons.

If one wishes to carry out such measurements on solids, the problem arises that the line width in the case of solids as a result of internal fields is considerably larger or the fall times of the excited nuclear magnetic resonance signals are considerably shorter, namely in the order of magnitude of only a few 10 s. Because of this, it is not technically possible to switch the required reading gradients or phase-coding gradients on and off during this extremely short signal duration, especially when strong gradients are required.

For this reason, various tricks have been used in the previously known isolated attempts to carry out volume-selective measurements on solids in order to reduce the line width of the solid-state signals or to extend the decay time of the signals.

In a known method of the type mentioned at the outset, a multipulse sequence (MREV-8) is used in order to achieve a narrowing of the lines. The multipulse sequence is also designed such that the magnetization is "stored" for a time sufficient to switch off the gradient magnetic field pulses.

However, the known method has the disadvantage that reading gradients have to be used for signal recording. However, these reading gradients cause a deterioration in the homogeneity of the constant magnetic field and thus a systematic line broadening. The known method is therefore limited to use with samples with narrow lines and does not permit any analysis of the line shape because the measured line shape is widened compared to the pure line shape due to the reading gradient for the reasons mentioned.

It is also known to counter the problem of wide lines in solid state measurements or the rapid drop in signal by rotating the sample by the so-called "magic angle". The rotation must, however, be carried out with a relatively high rotation frequency of several kHz, so that the rapid molecular movement of liquids can be simulated by rapid rotation of the solids and so that the local fields can be averaged. At such high rotation frequencies, however, the sample must be rotationally symmetrical for mechanical reasons. Even with rotationally symmetrical samples, however, there is a risk that the sample will deform due to the very high rotational frequencies.

This known method is therefore not suitable for measurements on biological samples, for example on extracted teeth, because such natural samples have an arbitrary, irregular shape and deformation of the sample cannot be permitted if, for example, a biological sample is still required after the measurement . This is the case in modern dentistry, for example, when a sick tooth is first extracted, then measured, then treated and finally re-implanted.

The invention is therefore based on the object of developing a method of the type mentioned at the outset such that two-dimensional and three-dimensional measurements are possible without the aid of reading gradients, so that line shape analyzes are also possible in the case of volume-selective measurements on solid samples.

This object is achieved in a two-dimensional measurement by first converting the magnetization of only one slice of the sample into a transverse magnetization tilted by 90 ° in a first time interval, and then switching on phase-encoding gradient magnetic field pulses in a second time interval that in a third time interval during the duration of the phase-encoding gradient magnetic field pulses, high-frequency pulses are irradiated onto the sample, which convert the transverse magnetization to be measured in the volume range into a multipole state which is insensitive to the phase-encoding gradient magnetic field pulses and up to over the time of the phase-encoding gradient magnetic field pulses being switched off in a fourth time interval, and that finally a high-frequency pulse is irradiated onto the sample in a fifth time interval, as a result of which the multipole -Status returned to a transverse magnetization and this is read out as a signal.

The object on which the invention is based is further achieved according to the invention in a three-dimensional measurement in that phase-encoding gradient magnetic field pulses are switched on in a first time interval, in that in a second time interval while the phase-encoding is ongoing Gradient magnetic field pulses radiofrequency pulses are irradiated onto the sample, which convert the magnetization to be measured in the volume range into a multipole state, which is insensitive to the phase-encoding gradient magnetic field pulses and up to the point in time that takes place in a fourth time interval Switching off of the phase-encoding gradient magnetic field pulses continues, and that finally a radio-frequency pulse is radiated onto the sample in a fifth time interval, as a result of which the multipole state is returned and read out as a signal.

The object underlying the invention is completely achieved in this way. By irradiating the high-frequency pulses during the duration of the volume-selective gradient magnetic field pulses, the magnetization of interest is converted into a state (dipole or quadrupole state) in which it is no longer sensitive to the magnetic field pulses and is long enough continues so that the magnetic field pulses can be switched off in a technically feasible manner. The subsequent reading out of the signals then takes place without the use of reading gradients, so that the line shape of the signals remains unadulterated and can be evaluated.

In this way, completely new volume-selective measurements can be carried out on solids. Not only can biological samples, for example bones or the previously mentioned extracted teeth, advantageously be measured, but measurements can also be made, for example, on plastics in order to determine, for example, the chain orientation of the plastics. If, for example, plastic is extruded in order to produce containers or the like, volume-selective measurement of the workpieces can be used to determine the molecular or chain orientation of the plastic as a result of the extrusion at various points on the workpiece. Since the chain orientation of the plastics is an essential measure of their strength, it can be determined, for example, in containers (plastic bottles) whether they are sufficiently stable in all areas.

In a particularly preferred embodiment of the method according to the invention for two-dimensional measurement, the transverse magnetization is generated in the first time interval by a first pulse sequence, which consists of a first, hard 90 ° high-frequency pulse, an immediately following spin lock pulse at 90 ° compared to the first There are 90 ° high-frequency pulses with a shifted phase position and a second 90 ° high-frequency pulse with a phase position shifted by 180 ° with respect to the first 90 ° high-frequency pulse.

This pulse sequence for generating the transverse magnetization, which is known per se from the document "J. Magn. Res. 83 , pages 299 to 308 (1989), has the advantage that after the completion of this pulse sequence the entire magnetization of the selected layer in z- Direction is aligned and "curled", ie not dephased, while the magnetization of the entire remaining volume outside the layer is uncontrolled dephased and therefore does not provide any signal components during the further course of the measurement.

It is furthermore particularly preferred in the two methods mentioned at the outset if the second pulse sequence radiated into and out of the multipole state consists of a 90 ° high-frequency pulse, one at a time interval there is the following first 45 ° high-frequency pulse and a second 45 ° high-frequency pulse which follows at a further time interval, the 45 ° high-frequency pulses preferably in phase relationship with the 90 ° high-frequency pulse alternatingly in phase or both around 90 ° shifted.

The use of this pulse sequence, known per se from US-Z-Physical Review 157 (1967), pages 232 to 240, has the advantage that the fact that it is converted into a multipole state is used to magnetize against the deactivation of the volume-selective gradients -Make magnetic field impulses insensitive and at the same time prolong the magnetization of interest so far that the magnetic field pulses can be switched off in sufficient time. The signal can then be read out without the presence of magnetic field pulses, so that the line shape is preserved in an unadulterated manner.

It goes without saying that the present invention, even if it is explained below using the example of proton nuclear magnetic resonance, can also be used for other types of magnetic resonance (e.g. ESR).

Further advantages result from the description and the attached drawing.

It goes without saying that the features mentioned above and those yet to be explained below can be used not only in the combination indicated in each case, but also in other combinations or on their own without departing from the scope of the present invention.

Fig. 1

eine Impulsfolge für ein erstes Ausführungsbeispiel des erfindungsgemäßen Verfahrens zur zweidimensionalen Messung;

Fig. 2

eine Darstellung, ähnlich Fig. 1, jedoch für eine dreidimensionale Messung;

Fig. 3

das Ergebnis einer bildgebenden Messung an einer Probe mit unterschiedlichen Materialbereichen;

Fig. 4

eine Variante zur Messung der Fig. 3;

Fig. 5

eine volumenselektive Messung von Kernresonanzspektren an einer Festkörperprobe mit unterschiedlichen Materialbereichen;

Fig. 6

vergrößerte Darstellungen zweier Kernresonanzspektren aus Fig. 5;

Fig. 7

zwei Protonenspektren einer Kunststoffprobe, die in unterschiedlichen Orientierung der Probe zum Magnetfeld aufgenommen wurden;

Fig. 8

ein Diagramm zur Veranschaulichung der Winkelabhängigkeit der Frequenz des Signalmaximums im Hinblick auf die Messung der Fig. 7.

Exemplary embodiments of the invention are shown in the drawing and are explained in more detail in the following description. Show it:

Fig. 1

a pulse train for a first embodiment of the inventive method for two-dimensional measurement;

Fig. 2

a representation, similar to Figure 1, but for a three-dimensional measurement;

Fig. 3

the result of an imaging measurement on a sample with different material areas;

Fig. 4

a variant for measuring Fig. 3;

Fig. 5

a volume-selective measurement of nuclear magnetic resonance spectra on a solid sample with different material areas;

Fig. 6

enlarged representations of two nuclear magnetic resonance spectra from FIG. 5;

Fig. 7

two proton spectra of a plastic sample, which were recorded in different orientations of the sample to the magnetic field;

Fig. 8

7 shows a diagram to illustrate the angular dependence of the frequency of the signal maximum with regard to the measurement of FIG. 7.

In the representation of FIG. 1, different high-frequency pulses are shown over time t among one another on an axis RF and gradient magnetic field pulses on axes G _x , _Gy and G _z .

For the two-dimensional volume-selective measurement of nuclear magnetic resonance signals on solid samples, a slice selection is first carried out with the aid of a pulse sequence 10, which consists of a first 90 ° pulse 11, an immediately following spinlock pulse 12 and a second 90 ° pulse 13, the entire pulse train 10 is radiated in the presence of a gradient magnetic field pulse 14 in the z direction.

With the first "hard" 90 ° pulse, the pulse train 10 folds the entire magnetization of the disk or layer defined by the gradient magnetic field pulse 14 into the z-axis and in this way generates a transverse magnetization.

The subsequent spinlock pulse 12 is phase-shifted by 90 ° with respect to the first 90 ° pulse 11, so that the high-frequency field points correctly in the direction of magnetization and is only effective for the spins that are in resonance, i.e. for those spins that are arranged in the selected disc. In this way, the magnetization of the disk is "lured", i.e. it does not dephase and is held in its state. The magnetization of the remaining volume areas of the sample, on the other hand, dephases in an uncontrolled manner so that it cannot contribute to further signal generation.

The pulse train 10, which is also described as a LOSY pulse train in the above-mentioned article in J. Magn. Res. 83 (1989), has the advantage that no shaped high-frequency pulses (sinc pulses or the like) are required, and that the z-gradient magnetic field pulse 14 is limited to an order of magnitude that only corresponds to the line width.

After the time interval τ₁ determined by the pulse train 10, a second time interval τ₂ follows, in which two phase gradient magnetic field pulses 15 and 16 are switched on in the x and y directions. These two magnetic field pulses 15, 16 are, as indicated in Fig. 1, phase encoding, i.e. their amplitude is switched through step by step.

It can also be clearly seen in FIG. 1 that a certain time is required to switch the magnetic field pulses 14, 15, 16 on and off, the duration of which is of considerable length in the context of interest here.

After the phase gradient magnetic field pulses 15, 16 are switched on, a third time interval τ₃ follows, at the beginning of a 90 ° high-frequency pulse 17 and at the end of a 45 ° high-frequency pulse 18 radiated onto the sample become. In this way, depending on the measured sample, a dipole or quadrupole state is generated, which on the one hand acts for a longer period of time than would correspond to the natural line width of a solid-state signal, but on the other hand also causes the signal in this state to be insensitive to the gradients -Magnetic field pulses 15 and 16.

It is therefore possible to switch off the two phase gradient magnetic field pulses 15 and 16 again in a subsequent fourth time interval τ₄, as can be clearly seen in FIG. 1, without the desired information, ie the magnetization, by premature decay is lost.

After the phase gradient magnetic field pulses 15 and 16 are switched off, in a fifth time interval τ₅, which again has the length τ₃, the signal can be extracted from the dipole or quadrupole by a second 45 ° high-frequency pulse 19. State can be traced back and read out as signal 20.

The time of the signal reading is thus after the end of the phase gradient magnetic field pulses 15 and 16, so that the line shape of the measured nuclear magnetic resonance signals is unchanged since there is no deterioration in the homogeneity of the magnetic field due to additional magnetic field pulses.

Within the scope of the measurement described above, two individual measurements are made for each amplitude setting of the phase gradient magnetic field pulses 15 and 16. This is done in that the two 45 ° high-frequency pulses 18 and 19 are both set in phase with the previous 90 ° high-frequency pulse 17 during the first individual measurement, while in the subsequent individual measurement the phase of the two 45 ° high-frequency pulses 18 and 19 is rotated by 90 °. This procedure, which can also be called serial quadrature detection, avoids the occurrence of mirrors during the signal acquisition. In this way, the total measurement time is doubled, but this is less critical for solid samples than for measurements on liquids, in particular on living human tissue, because in the latter case interference with artifacts must be feared in the case of long measurement times.

Through the two successive individual measurements with the phases of the 45 ° high-frequency pulses 18 and 19 rotated by 90 ° the real part and the imaginary part are measured in the time domain and thus the entire spectral information of interest.

The duration of the time interval τ₃ is set to achieve an optimal signal amplitude so that it corresponds approximately to the steepest drop in the free induction signal. In this way, approximately 56% of the initial magnetization is evaluated as an echo signal.

In the variant shown in FIG. 2, three-dimensional measurements are carried out.

In the case of FIG. 2, disc excitation is dispensed with at the beginning of the experiment and three phase gradient magnetic field pulses 30, 31, 32 in the x, y and z directions are switched on immediately, as is clear in FIG. 2 can be seen. If the three magnetic field pulses 30 to 32 are switched on, the experiment takes place in a similar manner as in Fig. 1 from the third time interval τ₃. For this purpose, a 90 ° high-frequency pulse 33 and then a first 45 ° high-frequency pulse 34 are radiated in at a time interval τ₃ in the experiment in FIG. 2, the physical phenomena being the same as explained above for the experiment in FIG. 1 were.

The three gradient magnetic field pulses 30 to 32 are now switched off and then after the time period t₄ a second 45 ° high-frequency pulse 35 is radiated onto the sample in order to subsequently read out a signal 36.

In this case too, the phases of the two 45 ° high-frequency pulses 34 and 35 are switched back and forth together by 90 ° in two successive partial measurements.

Some measurement results of the method illustrated in FIG. 1 will be explained below. The measurements were carried out on a tomograph with a superconducting magnet system with a field strength of 4.7 T and a proton measuring frequency of 200 MHz. Tubular gradient coils 30 and 15 cm in diameter were used. The maximum phase gradients were 7.5 G / cm. The gradient switching time was 2 ms. The phase gradients were adjusted in 32 steps.

For the measurements described below, a special probe head was used, the high-frequency solinoid coil of which had a diameter of 1 cm. The length of the 90 ° high-frequency pulses was 5 »s. The phase coding time τ₃ was set at 85 »s. The switching time of the gradient magnetic field pulses was about 2 »s, so that the time τ₄ for the transition to the multipole state was set at about 10 ms.

The measurements were carried out on samples consisting of hexamenthylbenzene and polytetrafluoroethylene (Teflon). At room temperature and a measuring frequency of 200 MHz, the spin-lattice relaxation time of hexamethylbenzene is about 360 ms. The dipolar relaxation time can be estimated at less than 100 ms, while the transversal relaxation time is about 40 »s.

3 and 4 each show the results of imaging measurements.

3a shows a first sample 40, which consists of a first material area 41, an empty space 42 and a second material area 43. The associated image recording is designated 44 and the two material areas 41 and 43 can be clearly recognized as white spots in the otherwise black background. Due to the 32 phase steps chosen, 32 x 32 phase coding with a corresponding number of pixels could be achieved. This resulted in a spatial resolution of 1.7 mm. Only the spin echo amplitudes were evaluated. The total measuring time was approximately 4 minutes.

A corresponding measurement is shown in FIG. 4 with a second sample 45, which had a total of three material areas 46, 48 and 50 and two empty spaces 47. As the image recording 51 shows, this finely structured sample 45 is also shown clearly resolved.

5 shows a third sample 60 with a first material area 61 made of hexamethylbenzene, a second material area 62 made of tetrafluoroethylene and a third material area 63 made of polyethylene.

As indicated by 61 ', 62' and 63 ', volume-selective nuclear magnetic resonances were excited in these three areas 61, 62, 63 and the corresponding nuclear magnetic resonance lines measured, as shown in Fig. 5 bottom right. There representations a), b) and c) show three spectral lines 61˝, 62˝ and 63˝, which belong to the three material areas 61, 62 and 63 mentioned at positions 61 ', 62' and 63 '.

It can be clearly seen from representations 5a), b) and c) that different signal intensities and different line shapes were measured. This is particularly clear from the enlarged representation of FIG. 6, where the lines 61˝ for hexamethylbenzene and 63˝ for polyethylene are shown again in detail.

The measured signal lines can now be evaluated from different points of view. For example, the associated frequency can be determined for each signal maximum, or the line width or higher-order moments. Provided that these parameters can be assigned to certain material properties, it is in principle possible to display these measurement results in an imaging manner by assigning an associated value on a predetermined gray scale to each measurement value.

FIG. 7 shows the measurement result of a sample of a polyparaaromatic amide, as is commercially available under the name Kevlar. During the measurements, the sample was aligned to the magnetic field in different ways, namely once parallel and once perpendicular to the magnetic field, as indicated by corresponding symbols on the two lines in FIG. 7. It can be seen that both the line shape and the position of the maximum change depending on the orientation to the external magnetic field. The angular dependence of the signal maximum is finally plotted in FIG. 8. The parameter shown here can e.g. are used to create contrasts in images in which the chain orientation of plastics is to be represented.

Claims (5)

1. A method for the two-dimensional measurement of magnetic resonance in defined small volume areas of a solid-state sample (40; 45; 60), the sample (40; 45; 60) being arranged within a constant, homogeneous magnetic field, being, further, irradiated in a predetermined manner with a sequence of high-frequency pulses (11, 12, 13, 17, 18, 19), and being exposed to a sequence of gradient magnetic field pulses (14 - 16) for maintaining during a period of time the spin magnetization to be measured, the period of time being longer than a switch-off time of the gradient magnetic field pulses (14 - 16), characterized in that initially within a first time interval (τ₁) the magnetization of one slice of the sample (40; 45; 60) only is transferred into a transverse magnetization tipped by 90°, that, subsequently, within a second time interval (τ₂) phase-coding gradient magnetic field pulses (15, 16) are switched on, that within a third time interval (τ₃) during the persistence of the phase-coding gradient magnetic field pulses (15, 16) high-frequency pulses (17, 18) are irradiated on the sample (40; 45; 60) transferring the transverse magnetization to be measured within a volume area into a multi-pole state being insensitive for the phase-coding gradient magnetic field pulses (15, 16) and being persistent beyond the moment in time of the switching-off of the phase-coding gradient magnetic field pulses (15, 16) within a fourth time interval, and that, finally, within a fifth time interval (τ₃) a high-frequency pulse (19) is irradiated on the sample (40; 45; 60), whereby the multi-pole state is again re-transferred into a transverse magnetization, the latter being read-out as a signal (20).
2. The method of claim 1, characterized in that the transverse magnetization is generated within the first time interval (τ₁) by means of a first sequence of pulses (10), consisting of a first, hard 90°-high-frequency-pulse (11), an immediately subsequent spin-lock pulse (12) having a phase being shifted by 90° with respect to the first 90°-high-frequency-pulse (11), as well as an immediately subsequent second 90°-high-frequency-pulse (13) having a phase being shifted by 180° with respect to the first 90°-high-frequency-pulse (11).
3. A method for the three-dimensional measurement of magnetic resonance in defined small volume areas of a solid-state sample, the sample being arranged within a constant, homogeneous magnetic field, being, further, irradiated in a predetermined manner with a sequence of high-frequency pulses (33 - 35), and being exposed to a sequence of gradient magnetic field pulses (30 - 32) for maintaining during a period of time the spin magnetization to be measured, the period of time being longer than a switch-off time of the gradient magnetic field pulses (30 - 32), characterized in that initially within a first time interval (τ₂) phase-coding gradient magnetic field pulses (30 - 32) are switched on, that in a second time interval (τ₃) during the persistence of the phase-coded gradient magnetic field pulses (30 - 32) high-frequency-pulses (33 and 34) are irradiated on the sample for transferring the magnetization to be measured within the volume area into a multi-pole state being insensitive for the phase-coding gradient magnetic field pulses (30 - 32) and being persistent beyond a moment in time of the switching-off of the phase-coding gradient magnetic field pulses (30 - 32) within a fourth time interval, and that, finally, within a fifth time interval (τ₃) a high-frequency-pulse (35) is irradiated on the sample, whereby the multi-pole state is again retransferred and is read out as a signal (36).
4. The method of one or more of claims 1 through 3, characterized in that the second sequence of pulses for the transfer into the multi-pole state consists of a first 90°-high-frequency-pulse (17; 33), a first 45°-high-frequency-pulse (18; 34) following after a distance in time (τ₃) as well of a second 45°-high-frequency-pulse (19; 35) following after a further distance in time (τ₄), the 45°-high-frequency-pulses (18, 19; 34, 35) having, preferably, a phase being alternately in phase and both shifted by 90°, respectively, with respect to the 90°-high-frequency-pulse (17; 33).
5. The method of one or more of claims 1 through 4, characterized in that the gradient magnetic field pulses (15, 16; 30 - 32) are set to be stepwise phase-coding.

Images